hsnsas style(anima)

This LoRA captures the illustration style of pixiv artist hsnsas (pixiv.net/users/58792218).

No trigger word required. The style activates automatically when the LoRA is loaded. Recommended weight: 0.7 ~ 1.0.

Tips:

Combine with character prompts for best results

Works well at any LoRA weight, adjust to your preference

Compatible with Anima base model

Training Details:

Base model: Anima base V1.0 (DiT)

42 images (832×1216 / 822×1200, aspect ratio bucketing enabled)

16 epochs

LoRA dim: 16, alpha: 16

Optimizer: AdamW8bit, LR: 1e-4

File size: 66.2 MB

本 LoRA 学习了 pixiv 画师 hsnsas 的插画风格。

无需触发词，加载 LoRA 后风格自动生效。推荐权重 0.7 ~ 1.0。

基础模型: Anima base V1.0

42 张素材（含 bucketing 自适应分辨率）

16 epoch 训练

LoRA dim: 16, alpha: 16

优化器: AdamW8bit, 学习率: 1e-4

文件大小: 66.2 MB

https://civitai.red/models/2809514/hsnsas-styleanima